1. Theory

  • Definition: A static induction transistor (SIT) is a vertical, short-channel, lightly doped JFET (or Schottky-gate MESFET) in which both gate and drain control a potential barrier in the channel, like the grid and plate of a vacuum triode.
  • Saddle-point barrier: Electrons must cross a potential barrier between closely spaced gate regions; its height is set by both gate and drain voltages.
  • Drain penetration: The channel is too short and narrow for the gate to screen it from the drain, so the drain field lowers the barrier directly.
  • Amplification factor: measures how much more strongly the gate controls the barrier than the drain does ().
  • Normally on or off: If the built-in depletion alone pinches off the channel (normally off); otherwise the device is normally on.

2. Working Principle

  • Barrier-controlled conduction: At low current, drain current rises exponentially as gate and drain voltages lower the barrier (thermionic passage over it).
  • Triode-like characteristics: The current keeps rising with instead of saturating, following an effective control voltage .
  • Space-charge-limited region: At higher current, the channel and drift region carry space-charge-limited current ( with ).
  • Voltage blocking: A negative gate bias raises the barrier so the device blocks about of drain voltage, up to gate-drain avalanche.
  • Bipolar mode: Forward-biasing the gate injects minority carriers that lower on-resistance by conductivity modulation, at the cost of slower switching.
  • Speed limit: The interleaved gate structure gives a large input capacitance, which limits .

3. Applications

  • Current status: Largely displaced by power MOSFETs and IGBTs in power electronics and by LDMOS and GaN HEMTs in RF.
  • Remaining niches: Some SiC SITs for pulsed power and high-power RF or induction heating.
  • Historical role: Linear high-fidelity audio output stages, high-frequency induction heating and broadcast RF transmitters.
NameFormulaRelated FormulaeVariables/Constants
Normally-on / normally-off criterion


Saddle-point barrier height

Barrier-controlled drain current



Voltage amplification factor

Effective control voltage
Triode-like region (empirical)

Blocking gain and blocking voltage
Specific on-resistance (unipolar limit)
Cutoff frequency

Important Questions

Ques: Why is the SIT’s output characteristic triode-like, while a JFET’s saturates?

In a long-channel JFET the gate depletion region shields the source end of the channel from the drain voltage, so beyond pinch-off () extra drain voltage appears only across a small depleted segment and the current stays nearly constant (see the Junction Field-Effect Transistor page). In a SIT the channel is short and narrow, so the gate cannot fully screen the drain: the drain’s electric field penetrates to the saddle point and lowers the barrier directly, which is the term in . The current then keeps increasing with exactly like a vacuum triode’s plate current, with replacing the triode grid-plate combination, and the sharpness of this effect is measured by the amplification factor . Reference: J. Nishizawa, T. Terasaki, and J. Shibata, “Field-effect transistor versus analog transistor (static induction transistor),” IEEE Trans. Electron Devices ED-22, 185 (1975); S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: How does a SIT block high voltage, and what does the blocking gain mean?

Making the channel narrow and lightly doped so that means the built-in depletion alone cuts off the channel, so the device blocks at zero gate bias (normally off). Applying a more negative gate voltage raises the barrier further, and since the drain lowers it only by per volt, the drain voltage that can be held off before the barrier collapses is about . The blocking gain, the ratio of blocked drain voltage to gate voltage magnitude, therefore tends toward , which is why a larger means a lower gate voltage is needed to block a given drain voltage. The practical ceiling is gate-drain avalanche breakdown. Reference: B.J. Baliga, Fundamentals of Power Semiconductor Devices, Springer, 2008.

Ques: What are the trade-offs that limited the SIT’s use in practice?

A normally-off SIT conducts only when the gate is brought near zero or slightly forward biased, and its channel is narrow, so its on-resistance is relatively high; operating it with a forward-biased gate (the bipolar-mode SIT) injects minority carriers that cause conductivity modulation, lowering in the same way as in a PIN diode, but at the price of stored charge and slower switching. Even in unipolar mode, the closely spaced gate structure produces a large input capacitance , which limits below the transit-time estimate . Together with the extra process complexity of a buried-gate vertical structure, this made the SIT a specialist power and high-frequency device, largely displaced in power electronics by MOSFETs and IGBTs. Reference: B.J. Baliga, Fundamentals of Power Semiconductor Devices, Springer, 2008.

Related pages in this vault: Metal Semiconductor Field Effect Transistor (MESFET), Junction Field Effect Transistor (JFET), Power MOSFET (VDMOS, LDMOS, Superjunction, SiC, Ga2O3, Diamond), GaN Power Diode, Schottky Diode, Long Channel MOSFET, Dual-Gate MOSFET, SiC Schottky Diode, Photodetectors (PIN and Avalanche Photodiodes).